University Place

The Big Picture: Astronomical Surveys

– Welcome everyone to Wednesday Nite @ the Lab,
I’m Tom Zinnen, I work here
at the UW-Madison Biotechnology Center.
I also work for the Division of Extension,
and on behalf of those folks and our other core organizers,
Wisconsin Public Television,
the Wisconsin Alumni Association,
and the UW-Madison Science Alliance,
thanks again for coming to Wednesday Nite @ the Lab.
We do this every Wednesday night, 50 times a year.
Tonight, it’s my pleasure to introduce to you two people.
Keith Bechtol and Ellen Bechtol.
Keith is in the Department of Physics,
and Ellen works for the Large Synoptic Survey Telescope
and also for IceCube.
Ellen was born in Charlottesville, Virginia,
and went to high school at Arlington, Virginia
at Washington-Lee High School.
She went to undergrad at the College of William & Mary,
in Williamsburg, Virginia and studied American studies,
and then got her master’s degree
at John F. Kennedy University in Berkeley, California
in museum studies.
Keith was born in Alexandria, Virginia
and went to Thomas Jefferson High School there.
He also went to the College of William & Mary
and studied physics, and went to Stanford
and got his master’s and PhD in physics there.
He did post-docs at the University of Chicago,
and here at UW-Madison.
Then he moved to Tucson, Arizona to be a staff scientist
with the Large Synoptic Survey Telescope,
and then in August, moved back to UW-Madison
to be on the faculty here in physics.
Tonight, they get to talk to us about the big picture,
science and public outreach with astronomical surveys.

Keith is going to talk first,
so why don’t I welcome Keith up to the lectern.
Please join me in welcoming Keith
to Wednesday Nite @ the Lab.
(audience applauds)

– All right, thanks a lot for having both of us here.
So this will be a presentation in two parts.
I’ll start by giving a little bit of an introduction
to astronomical surveys,
and then I’ll hand it over to Ellen and she’ll tell you about
all the education and public outreach opportunities.

So, this is a picture of the Large Synoptic Survey Telescope
that’s under construction in Chile,
so throughout this talk I’ll provide some examples
and point to LSST as an example
of one of these survey instruments,
but I want to actually start off by describing
what are astronomical surveys.
So, it’s a really exciting time to be giving this talk
and talking about astrophysics
because you may have heard in the news last week
that we were able to acquire the first image
of the event horizon of a black hole.
And this is a tremendous scientific breakthrough
and it’s also an example of how you can use
astronomical facilities to study
a particular object in detail.
So, on the right here is a radio image
of the black hole with the Event Horizon Telescope,
and on the left is an optical image
taken with the Hubble Space Telescope,
and these are extremely specialized observations
to achieve this high angular resolution.
The instruments that take this
are really designed specifically for this purpose.
Now, this is different from what we try to do
in survey astronomy. In survey astronomy,
we try to map as much of the sky as possible,
and it turns out that you want to use
different types of instruments for doing that type of research.
The reason is, is that even if you use an extremely powerful
and precise instrument like the Hubble Space Telescope,
you’re only able to image a small fraction
of the sky at a time.
So, the image here is taken by Hubble.
It’s one of the most famous images the Hubble has taken
where they looked at just what,
what appeared to be sort of a blank spot in the sky
and they see thousands of galaxies, right?
This image here, you can see the inset next to the full moon
and you see that it’s a really small fraction of the sky.
It would take something like 10 million photos of this size
to cover the full sky.
So Hubble is an extremely powerful instrument,
but it’s not the instrument that you would use
if you want to map as much of the universe as possible.
We think that there’s something like two trillion galaxies
in the whole observable universe, and right now,
we’ve seen just a fraction of them.

So the idea of an astronomical survey
is that you try to cover as much of the sky as possible.
You go very wide, you try to image as quickly as possible
so you can scan the sky,
and you try to go as deep as possible
so you see as many astronomical objects as possible.
So this is showing
the field of view of the Dark Energy Camera,
which is one of the instruments that I’ve used,
and you can see that it’s actually much larger
than the full moon, right?
This is a field of view that’s substantially larger
and actually gives us a reasonable chance
of mapping large portions of the sky.

So we’ve been going on this enterprise now for several years
and this is showing a representative patch of sky
that corresponds to half of one of the individual CCDs
on the previous image.
If you look closely here,
you can see that the whole focal point of the Dark Energy Camera
actually consists of 62 individual CCDs, arranged in a mosaic.
So if we could just zoom in on one half
of one of those CCD chips,
every single one of the objects encircled in green
represents an individual star or galaxy
that’s been detected in the images.
And so, by mapping approximately 1/8 of the sky
over the first 300 nights or so of operations
of the Dark Energy Camera,
we’ve now cataloged something like
400 million individual stars and galaxies,
and these are going out to objects that are
10 million times fainter than we can see with the human eye,
galaxies where we’re seeing light that was emitted
10 billion years ago.

This is another glimpse of what survey astronomy can do.
This is a composite image assembled with data
from the Gaia spacecraft,
which was launched by the European Space Agency.
So in this image, you see the collective light
of the brightest two billion stars
that are visible from Earth right there.
They’re mapping the entire sky looking at
the very brightest stars,
and so you can create beautiful images,
not just of a portion of the sky,
but really getting a synoptic view of the entire universe
as from our perspective here on Earth.

One of the really amazing things about taking these images
is that we’re actually sensitive to be able to see objects
and to do science on a huge range of different size scales.
So this is just one image with the Dark Energy Camera,
and I like this image because you can see physical phenomena
on just a huge range.
So let’s start from the, you know the most distant objects
and work our way back in.
If you could look closely at this image,
you would see the smallest smudges of light
are distant galaxies that are billions of light years away.

And these comprise most of the objects in this image.
Now these larger extended spiral galaxies
that you can see in the foreground, there’s two of them,
these are nearby galaxies.
These are maybe, you know, a few million light years away.
They’re close enough that we can actually see
the structure of the galaxy easily.

Now this bright blue point of light in the lower left,
that’s one of the stars in our own Milky Way,
so that’s something that’s– we’re light years away,
maybe hundreds of light years away,
so much, much closer.
And then it’s a little bit difficult to see,
but in the very top right-hand corner,
you can see an object that’s two neighboring points of light.
One is red, and one is green, okay?
We interpret that as an asteroid that was moving through
this field of view, and as we were taking pictures
of this part of the sky in different color filters,
it showed up in one color filter, you know on one image,
and then showed up in another color image a bit later.
And so this is an object that’s in our own solar system
that’s maybe light minutes or light hours away.
All right, so in this single image,
you see just this enormous range of astronomical phenomena,
and an enormous range of science that you can do
with one dataset.
So why astronomical surveys?
Why are we so interested in this?
Well so I’m a physicist,
and so the way that I like to think about this
is that our universe is just one realization
of a statistical process
that’s governed by the rules of nature.
And so we kind of get one opportunity
to see how these laws of nature have built up
all the galaxies in the universe that we can see.
As a physicist, we aim to discover what these rules are.
And the interesting thing is that some of the patterns,
some of the physical laws that we would like to discover,
there are effects that only appear when we look
at the universe on the very largest scales.
They’re not something that you can see
by looking at a single star or a single galaxy.

To give a sense of what I mean,
I want to show an image of the cosmic microwave background.
So this is not optical light, but looking at
a part of the radio or the microwave spectrum.
This is taken with the Planck spacecraft
and it’s showing the color scale here
is representing density fluctuations in the very early universe.
This was light that was emitted
about 400,000 years after the Big Bang.
And the density contrast between the least dense
and the most dense regions of the universe at that time
was about one part in 100,000.
That’s about the same level of fluctuations in density
as when you’re speaking, the density of air
that’s created by your voice, okay?
So very, very minute perturbations.
Out of that grew all of the galaxies
that we have in the universe today,
and when we look at an image like this,
the way that we characterize it is statistically.
All right, we look at the pattern of the denser spots
and under-dense spots,
and this reminds us that what we’re actually seeing
is literally quantum fluctuations
that have been blown up to the scale of the universe, okay?
Perhaps the most amazing thing about all of this
is that at least on one planet,
there’s intelligent life that’s, you know,
looking out there and trying to understand what it means,
and our place in it.
And the remarkable thing is that the universe
seems to be governed by mathematical laws.
We actually have some chance of understanding it,
and understanding what our context is in this.

One of the really surprising things that we’ve learned
over the past couple decades
is that even with our powerful optical telescopes
we’re only seeing a really small fraction
of all the matter and energy that’s out there
in the universe.
It turns out that something like 95%
of all the mass and energy in the universe
exists in forms that are invisible to us.
We give these phenomena names like dark energy,
dark matter, neutrinos.
We know that there are distinct physical processes,
but in the large sense, these names are placeholders
for physics that we don’t yet fully understand.

And what’s, you know, kind of perplexing
and makes us so curious about these
is that we can’t touch them, we can’t feel them,
they’re invisible and yet they determine the fate
of the universe, right?
So we would like to understand them.

Let’s focus on one of these just to go a bit deeper.

So one of the topics that I study is dark energy.
This seems like it’s some form of energy
that permeates all of space and is constant in time.
It has the effect of driving galaxies apart,
so over time, the spaces between galaxies are growing
at an accelerating rate.
And the easiest way for us to explain that
is that there is some form of energy density
out there that’s driving them apart.
Now, you might think, “Okay, there’s all this energy out there.
“Why can’t we sense this? “Why can’t we see this?”
So it turns out that if you use Einstein’s famous formula
of relating energy to mass,
if you were to take this dark energy and ask,
“What is that corresponding density
“in terms of matter particles?”
It turns out that the density
is about four protons per cubic meter.
And this is an energy density that’s everywhere.
It’s in this room, it’s in the space between stars.
It’s in the space between galaxies.
It’s what’s driving the overall evolution of the universe right now.
It’s the largest component.
Now, if you compare that extremely small density,
four protons per cubic meter to say
the density of interstellar space, right.
Interstellar space, we think of it as a vacuum,
but it’s something like a million protons per cubic meter.
The air in this room is something like 10 to the 27,
that’s a one with 27 zeros after it,
protons per cubic meter.
And something like water like what our body is made of
is 10 to the 30.
So in terms of what we can sense here on the earth,
it’s extremely small, but the universe is really big,
and so when you average over the entire universe,
this dark energy is actually by far the dominant component,
and that’s pretty surprising.

So you might wonder, “How can you study something like this
“that’s invisible, that you can’t touch it,
“you can’t feel it?”
And the answer is that we have to look at
statistical properties of the universe.

And this is where the astronomical surveys come in.

So this is a map of galaxies that’s showing
one narrow slice in a large,
three-dimensional map of galaxies.
And each of the yellow points here
represents the location of one galaxy.
As you go further out from the center of the image,
you’re looking at galaxies that are further and further away
and you may notice if you look at this image for a moment
that the galaxies aren’t randomly distributed,
they’re actually sort of clustered into filaments
and kind of like bubbles and clusters and so forth, right?
And then there’s large gaps between them.
And this structure here is actually exactly
what you would predict from doing cosmological simulations
and understanding how gravity operates on matter
over billions of years.
So we can do precise numerical simulations
and we could compare the distribution of galaxies
that we would predict to that which we observe.
And again, this pattern of structure,
these filaments and these elongated structures,
it’s something that’s only visible on the scale
of hundreds of millions of light years,
or at least tens of millions of light years.
It’s not something that you could detect
by looking at a single galaxy.
It’s something that we need a large map of the universe
to be able to see.

Another example of one of the tools that we use
to study dark energy are supernovae.
Now these supernovae, they’re extremely luminous
stellar explosions that we can see
over great distances in the universe,
and they’re excellent tools for studying dark energy.
The challenge is that these supernovae
are extremely rare and unpredictable.
In a given galaxy, you may have one supernovae
per 1000 years, right?
But in order to study these supernovae in detail,
we like to acquire additional observations
with other telescopes,
and so how do you plan these observations
when you don’t know when the supernovae will occur?

The solution is that you just look
at as many galaxies as possible.
If you look at thousands and thousands of galaxies,
then in the course of a human lifetime,
in the course of an experiment, right?
You don’t have to wait forever,
you’ll see some of these supernovae,
and the people who pioneered this called this
supernova on demand, ’cause they didn’t know where
the supernova would occur, but they knew if they looked
at thousands of supernova that they wouldn’t have to wait for too long.

So once you have one supernova, that’s not enough, right?
You actually want hundreds or thousands,
or even hundreds of thousands of supernova
to actually see the cosmic expansion history and so this is
a plot from one of our recent papers,
and you don’t need to understand the details,
the point of this is just that by collecting the data
from many, many, many supernovae,
you’re able to distinguish between different models of dark energy,
but you couldn’t so that with only a single supernova.
You need this ensemble of measurements
in order to make the inference.

Another technique that we use is called
weak gravitational lensing.
This is a way of measuring
the total mass distribution in the universe.
And so this image is showing the mass distribution
over about 1/20 of the full sky.
The regions in red and blue are showing regions
that are either more dense or less dense.
And again, this pattern of seeing regions
that are more dense or less dense,
it’s only a pattern that you could see
by making this large map of the universe.
If you only looked at a single galaxy,
you wouldn’t even be able to measure
this weak gravitational lensing effect.
It’s an example of something that’s
entirely a statistical measurement.

One more example that’s a bit closer to home.
So far, we are talking about, you know,
the whole distribution of galaxies throughout the universe.
What if we just talk about our own Milky Way?
Our own Milky Way galaxy that’s hundred billion stars or so.
Well if you actually look at those stars in the Milky Way,
you see that they are also not randomly distributed,
but they actually have many structures
around the Milky Way,
and in particular, some of the most dramatic structures
are these so-called stellar streams.
These are like rivers of stars in the sky
where another galaxy has come too close to the Milky Way
and it’s been pulled apart by the gravitational attraction of our own galaxy.
When it gets pulled apart, it forms these elongated streams of stars.
And again, this is something that you couldn’t see
if you only looked at individual stars,
but we have maps of now something like 1/8 of the sky,
very detailed digital maps,
and so we can measure the distances of different stars
and we start seeing these structures.
So on the left-hand panel here,
this is showing a map of the stellar density
and it’s a movie that’s stepping out in distance,
it’s scanning in distance.
And at particular slices in distance,
you may see these faint black streaks
that are crossing the image.
These are the so-called stellar streams, right?

Where these regions where the stars
are all moving coherently, and telling us something about
how our own galaxy formed.
This is something that you wouldn’t be able to access,
except by making these large maps of the sky.

So let’s talk about how we do this.

This will transition into a bit of the technology
of how this is possible.
So over the past couple decades,
there’s been a progression in our ability to map the sky
faster and deeper and wider.
And so one of the landmark players in this game
was the Sloan Digital Sky Survey,
which started in the year 2000 and is still going today.
The University of Wisconsin is a major player
in the Sloan Digital Sky Survey.
This was an incredible new instrument.
It was using a relatively modest telescope,
only 2.5 meters across.
I say “only” but that’s actually fairly modest
in the landscape of professional astronomy.
But it is arguably the most productive scientific instrument
in astronomy in all of its history.
What it did is that it created
the first digitized maps of the night sky
and it made all of its data public,
so that thousands of scientists
could all analyze this data.
When you go to an astronomy colloquium and someone asks,
“Hey, who here has used SDSS data, the Sloan data?”
Basically everyone in the audience
will raise their hand, right?
It changed the landscape of astronomy,
and things would never be the same after SDSS.
Now the project that I work on now is called
the Dark Energy Survey, or DES.
That started in 2012,
and that’s about 10 times deeper than SDSS,
so we can see galaxies and stars that are much, much fainter.
And the project that we’re building in Chile
is called the Large Synoptic Survey Telescope
and that’ll start running in about 2023 or so,
and run for about a decade.
LSST, as it’s known, will image stars and galaxies
about a hundred times fainter than those
that were detectable by SDSS.
And so you can see that there’s been
this step forward in technology
that’s allowed us to map the sky with ever more precision.

Now the big difference between these different surveys
is their amount of light-collecting power.
So the word that we use for this is “etendue.”
And you can think about it as basically
how many photons from the night sky
that you’re able to collect with your telescope.
The etendue is the product of the field of view,
basically how wide on the sky you can image at once,
the effective aperture, so the size of your mirror,
and your operating efficiency.
And this plot here is showing a number of different surveys.
The vertical axis is a log scale,
and so you can see that LSST is something like
a factor of 10 more in its light-collecting power
than any previous instrument, okay?

So how is this possible?
What it required is actually over two decades of planning
leading up to the moment that we have today.
LSST is really an entirely new kind of telescope
that’s been optimized specifically for surveys.
The first discussion of LSST,
you can trace back to around 1998,
when scientists realized that there was this optical design
that would allow them to achieve this very large field of view.
In the year 2010, LSST was selected
by the whole astronomical community
as the highest priority project of the next decade,
and then the formal construction of LSST
started in the year 2014.
It’s a joint project between the National Science Foundation
and the Department of Energy, and our international partners
and that’s a theme, right? These projects are enormous.
LSST is something like a $700 million project.
But what it will do is create a dataset
for literally thousands of scientists
and even thousands more educational opportunities
for students and anyone who’s interested in astronomy.

So what made this possible
was a confluence of enabling technologies.
It really took all three of these
to make the LSST design feasible.
What we needed are extremely large mirrors,
so in particular, eight-meter glass mirrors
that are made at the University of Arizona.
It took a mosaic of high quantum efficiency CCDs
that are optimized for near-infrared sensitivity
so we can see the very most distant galaxies.
And importantly it took exponentially increasing computing power.
Again, the vertical axis on this plot is a log scale,
and each of the different types of symbols
is representing the growth in these different aspects
of the technology, and you’ll see, for example,
this computing power, this is growing with Moore’s Law.
The number of transistors that we can pack
is growing exponentially with time.

So let’s step through some of these ingredients.
The first main ingredient is having
this enormous field of view,
so you can see represented here,
the field of view of several different surveys.
LSST is in the lower center for reference,
and you can see the full moon there
on the lower left-hand corner.

The next ingredient that you need is a very large aperture.
So this was a picture taken of the glass
that’s used for the mirror of LSST,
and you see all the people who were there on this day
to give a sense for just how large this mirror is.
This was before it gets silvered with its reflective coating.
But this mirror is 8.4 meters across,
it’s about the size of this room, okay?
An enormous amount of light-collecting power.

The next key ingredient is a camera.
The LSST camera is about the size of a car.
It’s the largest, most powerful camera
ever built in astronomy.
To give a sense of the key numbers of the LSST camera,
it has 3.2 gigapixels; 3.2 billion pixels.
If you were to take a single image with LSST
and project it at its intrinsic resolution
using high definition TVs,
it would take the area of about a half a basketball court.

That is a huge amount of data, right?
And we take one of these images every 15 seconds for 10 years.
That’s the goal.

This is an image of the commissioning camera
showing the array of nine of the CCD sensors.
This commissioning camera we’re using as a prototype
to help us make sure that the different parts of LSST are functioning correctly.
And it’s using only one of 21 rafts
that will make the full LSST camera,
so you can see in the lower left-hand part of this image
someone’s hand to give a sense.
The full LSST focal plane is the size of a manhole cover.
And it’s all instrumented with silicon CCDs.

This is an image of the lens
that’s on the front of that camera,
compared to a human for scale.
Right, so you can see how enormous this is.

And we’re collecting something like
10 terabytes of data a night.
That’s enough to fill up multiple hard drives
of all your laptops.

We have fiber optic that’s going from the cable
and leads all the way to the supercomputing facility
at the University of Illinois.
There is a bank of a computers that are capable
of achieving processing speeds up to two petaflops.
Peta is 10 to the 15, okay,
operations per second.

And this is one of the big challenges
of data management, right?
Once we have all these images, we need to analyze them,
we need to extract all of the individual stars and galaxies
and measure their properties.
Now with SDSS shown in the lower left,
we had the advantage so to speak that the galaxies and the stars
were all kind of distinctly located on the sky.
At the LSST depth shown on the right,
we see so many stars and galaxies
that all of their light blends together.
It’s a major processing challenge
just to be able to distinguish the individual objects.
We see that the universe is just full of light,
full of galaxies.

So over 10 years, LSST aims to visit
every patch that’s visible from Chile,
about a thousand times.
The idea is that it will return to each patch
of the night sky about every three nights,
and so by collecting these thousands of images
over 10 years, we’ll actually not only be able
to get like one very, very deep image,
we’ll actually be able to see everything
that’s moving and changing over that 10-year period.
You can think of it as a 10-year color movie
of about 40 billion stars and galaxies,
and all the solar system objects as well.

Here are some of the LSST key numbers just to give
a sense of the magnitude of what we’re talking about
in terms of the sheer numbers.
So this is shown in a graphical representation,
but if you just were to pick out some of the key numbers from this chart,

these are some of the ones that I would highlight.
They have this 3.2 billion pixel image every 15 seconds,
365 nights a year for 10 years.
That’s about 5 1/2 million images.
About 500 petabytes of images, total,
collected over 10 years.
There’s about 10 million alerts that are distributed each night
and they’re released to the world within 60 seconds
of the shutter closing on the camera.
There’s about 20 billion stars and about 20 billion galaxies,
so LSST will actually catalog
something like 1% of all of the observable galaxies
in our universe.
There’s actually a finite number of galaxies
that we can see, because the universe is a finite age.
Like the light can only travel so far,
and so with LSST, we actually start getting
to an appreciable fraction of literally every galaxy
that we could conceivably see.
There’ll be something like 30 trillion
individual flux measurements.
So when you have an instrument like this
that actually changes the entire way
that you think about astrophysical research,
and the whole landscape of professional astronomy is changing.
Nowadays, we typically work in large collaborations
like the one shown here.
This is one of our meetings for the Dark Energy Survey,
where we have hundreds of collaborators
that are all working together to try to pull off
a survey of this challenge,
both technically and scientifically.

And another big thing that we’re doing
is that we’re making all of this data public.
So instead of this model where,
if you’re an astronomer,
you work with your particular grad students,
you write your proposal,
you go to the telescope and you take the data back on
one of your thumb drives,
now what you do is you write a SQL query.
You go to the database and you download
the particular objects of interest,
and whether you’re interested in solar system objects,
stars, galaxies,
it’s just making a different query, right?
The idea is that this dataset will be useful
for many, many different scientific purposes.
So you can think about all the science that you could do
with this archival data.

So, some of the values that we have in survey astronomy
are shown on this slide.
We increasingly are moving towards open source code.
We try to make as much of our data public as possible,
and by doing this, we ensure scientific reproducibility.
We try to make it possible for scientists
to basically run the exact same algorithms,
run over the exact same data and see if they get to the same result,
or maybe they’ll find a better way of doing it.
We emphasize collaboration,
and we really emphasize equal opportunity,
that by making this data public,
anyone with a good idea can try to find something new.
So I think this is a good point to hand it over to Ellen
to talk about education and public outreach opportunities.
(audience applauds)

– Hi. Okay, so I’m going to talk about survey astronomy
from a slightly different perspective,
that of education and public outreach.
So you just heard a lot about the exciting science
that can be done with survey astronomy,
and I feel like there’s equally exciting
and numerous opportunities to engage with survey astronomy
as a member of the public, or in educational classrooms.
So, I’ll use LSST as a case study,
but I wanted to start with a little bit of context
from a historical perspective.
So SDSS, the Sloan Digital Sky Survey,
was especially important scientifically
and also from an education and outreach standpoint.

And so they have a long history with bringing survey data
and even some other survey instrumentation and objects
itself into the classroom.
So this is a screenshot from their website, the Education section of it,
and you can see on the right-hand side there,
there are three people and the person in the middle
is holding this giant plate.
And that’s a plug plate that is used
as part of the survey to take their images,
and they distribute these plates to teachers
around the country, certainly, when they’re done using them.

Another project of SDSS
that really was quite wildly successful
was called Galaxy Zoo, and in this instance,
they made a subset of the data available
for the public to help them classify
which types of galaxies they were seeing.
There are so many galaxies,
they couldn’t do it all on their own,
one scientist in a room alone,
so they made the data publicly available.
And so in the first Galaxy Zoo,
there have been many additions since the original one
in the year 2007, but the first one, since then,
there have been more than 40 million classifications
and they were made in approximately 175 days
by more than 100,000 volunteers.
And they thought that it would take volunteers years
to complete that dataset,
and it didn’t take that long at all.

So this is not meant to be representative
of all of what SDSS or DES do for education and outreach,
but just to give you a flavor in a sense
for what some other surveys have done.
So the Dark Energy Survey,
they have several projects,
one of which is called Dark Bites, and this is really
a blending of art and science.
So they use several of their talented collaboration members
both to create artistic representations
and pair them with facts and analogies about the universe.
And then this is distributed online
and through social media and has thousands
to hundreds of thousands of views and impressions over time.

Another project is called Scientist of the Week
and they also have some reflections of the observing teams
and through this work, I’ve noticed that they’ve really
highlighted the diverse individuals,
the diverse skillsets and backgrounds
and areas of expertise that it takes
to make one of these surveys work, right?
It takes a lot of people around the country,
around the world with a lot of different skills
to make these telescopes go day and night.

They’ve also translated several other
education and public outreach websites
into as many as four languages.

Okay, so for LSST, education and public outreach,
our mission is to provide worldwide access to,
and context for, LSST data through accessible
and engaging online experiences
so anyone can explore the universe
and be part of the discovery process.

So among ground-based astronomy projects,
LSST is fairly unique in that we have substantial effort
and funding during the construction phase
to create education and public outreach
products and materials.
For a lot of the previous astronomical surveys
and science in general, education and public outreach
relies heavily on volunteers
to just donate their time and efforts and energy
to make education and outreach happen.
For LSST, there is dedicated staff and funding
in the construction phase to build up an EPO program
before the survey actually is running.

We have four main audiences:
Formal educators at the level of advanced middle school,
high school, and Astro101, or intro college-level courses,
citizen science principal investigators,
content developers at science centers and planetariums,
and the science-interested teens and adults,
such as yourself, the general public.

So during construction,
EPO will build an operations website
and materials in both English and Spanish,
a formal education program based on online notebooks,
a repository of multimedia resources,
an interactive Skyviewer, a cloud-based education
and public outreach data center,
the capability for researchers to build their own
citizen science projects using LSST data,
a communications and marketing plan,
and a strategy for measuring success,
which is what I personally work on,
the evaluation component.
So I’m going to step through some of these,
these products right now.
So for formal education, we are building themes,
investigations around six themes,
and these six themes are based heavily
on the type of science the LSST
will be really good at producing.
So they are Properties of Light, Properties of Stars,
Galaxies and the Milky Way, Cosmology,
the Solar System, and the Dynamic Sky.
And we’re building these investigations
in close collaboration with educators.
We have several educators on our staff.
And we’re also actively doing user-testing
and needs assessment with educators
to make sure that what we’re building is well-aligned
for their needs and their interests in the classroom.

So here’s an example of one of our current investigations.
Everything will be on a webpage,
so it’s easy to access, there’s low tech startup costs.
If you can load a website, you can load the investigation.
You don’t need to download any data,
you don’t need to download any special software
or anything from the database.
In this particular investigation is walking students through
how telescopes collect data and then how that data
is turned into astronomical imagery.
And we’re building an interactivity
so we will have a subset of LSST data and images available
where then students can apply different filters
and different amounts to make their own
astronomical images and then share them out
with their friends and family and their teachers.

For planetariums and science centers,
we are creating a library of digital multimedia assets
including a Fulldome sky view of that LSST alert stream
that Keith mentioned not that long ago
made available every 60 seconds.
I don’t think our Fulldome sky view will be made available
every 60 seconds, but it will be updated on a regular basis.
We are also making media specific for Chilean audiences
because the telescope is in Chile.
We have a strong education and public outreach component
in that country, and we’re also making 3D materials.
We are providing a multimedia gallery
which is actually available now,
and I’ll show several images from the gallery
at the end of this talk,
and you can see the website there, gallery.lsst.org.

And these assets will be provided in formats
that support emerging industry standards,
so Data2Dome, AVM, and IMERSA Dome Master.
So these are all standards that the planetarium
and science center community are using
and will make it easier for content creators
and developers at those institutions
to use the imagery however they see fit.

So the Skyviewer, this is specifically targeted
for members of the public,
and we see this as really a self-guided educational tool,
so there will be a website with some subset
of the LSST dataset available.
And we are hoping to curate and planning to curate
objects of interest and even make some guided tours,
so you can explore a topic,
or a theme of particular interest to you,
and we’ll take you around the LSST night sky.
And you can learn more about it and dig in
to all the different aspects that you’re interested in.

Everything will be mobile friendly.
You can use it on a tablet, on your cell phone,
on the laptop, including the Skyviewer,
and we’re planning to make as much of
the education and public outreach products as possible
available both in Spanish and in English.

So for Citizen Science, I highlighted previously
the Sloan Digital Sky Survey Galaxy Zoo project.
That project was so successful, it really launched a whole
team of people who run the Zooniverse, zooniverse.org,
it’s an online citizen science platform
that’s been very successful.
It has citizen science projects available
from all different fields, not just astronomy,
and we’re working with that group
to create a pathway so that LSST scientists
who have an idea
to create a citizen science project can do that
and make a subset of the LSST data available
for members of the public to help with scientific research.
So it’s another way to take authentic data
out to the public.

And here’s an example of Space Warps HSC, so this is
a version of the Space Warps project,
using Hyper Subprime-Cam survey data.
And a few months ago this was still running
and when I went a couple days ago
to get this screenshot, it’s already over.
Everything’s been classified.
And when that happens, you can still go online
and classify objects,
you’ll just get more of the test imagery.

So that’s really exciting that these projects continue
to be of interest to the public.

Okay, so I thought it’d be fun to show off
some of the nice imagery and photos
and highlight some of the recent construction
events and milestones with the project.
So here you can see this is the mirror coating chamber
and it is on its way to the summit.
In the bottom right-hand corner,
you can see that it was a very tight fit
going through a tunnel.
So this is the chamber that will provide the mirror
its both reflective and protective coating,
and that is done actually at the summit site.
And so this is actually half of the coating chamber,
it had to be split into two sections
in order to be transported.

It was transported initially from France.

And so here it is on its way, it’s nine meters in diameter
and it took a lot of logistics planning
and a number of different people to make this happen.
There were some lights that had to be removed,
and street signs, and cords that were hanging down
so nothing would get hung up.

Here’s a photo of the telescope mount assembly,
so this is actually what the camera will be mounted onto
and this is in Spain, so again,
you can get a scale,
a sense for how big of a project this is
and it takes people all around the world
with all different backgrounds
and areas of expertise to make this work and happen.

So here’s a picture of the primary tertiary mirror for LSST
in the foreground with the blue protective coating on top
and behind it is the mirror cell,
so that’s what the mirror will actually sit on
when it’s at the telescope site.

And in this picture,
this is at the Richard F. Caris Mirror Lab,
mirror facility in Tucson, Arizona
at the University of Arizona,
and in this picture, suction cups are being used
so it’s a vacuum method to pick up the mirror
and put it into its shipping container.
And there it is being shipped, so it’s on a highway
going to a port where it got loaded onto a boat
to then go via boat to Chile.

So that also give you a sense of scale.
And to close out the talk,
this is the most recent picture that I could find
of the telescope site, so this was taken back in March.
And you can see a couple cranes in there
if you look closely and a bunch of cars.

I hope it gives you a sense of the scale
and just general like, excitement
for working on a project like this.
So if you want, you can follow along
with the construction process at lsst.org.
You can see some of these images at gallery.lsst.org.
You can follow us on Facebook, and also Twitter.
And thank you for being here tonight.
We’ll take some questions.
(audience applauds)